I recently had a total Aha! moment (or a why-didn’t-I-ever-think-of-it moment) when I had chanced upon a recently published article titled “Eating Expectancies in Relation to Eating Disorder Recovery” by Fitzsimmons-Craft and colleagues. The title caught my attention because I had never come across any research tying eating expectancies to eating disorders, though I was familiar with the concept from the health psychology and obesity literature. Eating, as a behaviour and as a mechanism, is incredibly complex, with many factors contributing to why and how we eat; eating expectancies are one such factor.
Expectancy theory, first proposed by Tolman (1932), suggests that expectancies, or assumptions about the consequences of various behaviours, develop as a result of one’s learning history (Smith et al., 2007). Such expectancies are thought to influence subsequent behavioural choices, with one acting to either increase the likelihood of reward or decrease the likelihood of punishment. Essentially, expectancies are cognitive mechanisms that drive future behaviours.
With respect to eating, expectancies represent the culmination of one’s learning history as related to eating and act as primary cognitive mechanisms that guide one’s future eating behaviors (Fitzsimmons-Craft et al., 2013, p. 1041).
Eating expectancies can be negatively reinforcing, that is, eating helps alleviate a negative emotional state or relieves boredom. Eating expectancies can also be positively reinforcing: one, eating can be pleasurable and rewarding, and two, it can enhance cognitive competence (Simmons, Smith, & Hill, 2002).
EXPECTANCY THEORY: REFINING OUR UNDERSTANDING OF EATING DISORDER RISK FACTORS
Expectancy theory could be useful in helping us understand the literature on ED risk factors (Smith et al., 2007). For instance, both the perception of pressure to be thin and a history of being teased about weight are known predictors of eating disorder symptom endorsement (Stice, 2001; Gardner et al., 2000; Thompson et al., 1995). Both factors have also been shown to predict dieting and body dissatisfaction (e.g., Cooley & Toray, 2001).
Therefore, these experiences could contribute to “the expectancy that dieting and thinness lead to reinforcement” (Smith et al., 2007, p. 189). Rather than merely measuring the exposure to events that may influence one’s learning (e.g., a history of being teased about one’s weight), expectancy theory advocates that we should measure the degree of reinforcement one expects from, for example, thinness and dieting, and one’s beliefs about the consequences of eating.
In other words:
Expectancies are useful in understanding eating behavior because they assess actual learning and thus represent a proximal, presumably causal, risk factor in the eating disorder trajectory (Hayaki, 2009, p. 553).
Lastly, as opposed to emphasizing the actual reinforcement value of one’s behaviours, expectancy theory focuses on the individual’s cognitions.
EVIDENCE FOR THE ROLE OF EATING EXPECTANCY IN DISORDERED EATING
In their development of the Eating Expectancy Inventory, Holstein et al. (1998) discovered that the expectation that eating helps mitigate negative affect and boredom (i.e., expecting negative reinforcement from eating) characterized patients with bulimia nervosa (BN) but not patients with anorexia nervosa (AN) and psychiatric controls.
Bohon and colleagues conducted a year-long investigation of factors that predict the maintenance of bulimic pathology in a community sample of individuals with threshold or subthreshold BN (2009). They found that expecting reward from food intake, or eating expectancies related to positive reinforcement, was associated with a longer time to remission from binge eating. (Note: using a community sample increases our ability to generalise the findings; and conducting surveys repeatedly across a year increases our confidence in the results.)
Do initial binge eating and purging behaviours influence the subsequent trajectories of expectancy development?
Smith et al. (2007) wanted to determine whether, in a sample of middle school girls, initial binge eating and purging behaviours would predict, two years down the line, that these girls would experience an increase in the expectancy that eating reduces a negative emotional state. They found that those who engaged in binge eating behaviours were more likely to expect eating to decrease their negative state whereas those who did not binge eat were less likely to endorse that belief.
The results support the possibility that binge eating influences the development of eating expectancies:
The direction of the effects is noteworthy. The binge eaters tended to grow in the belief that eating helps alleviate negative affect, and low levels of binge eating were associated with a decline in that belief. There was no evidence of a corrective process in which binge eating led to reduced expectancies that eating helps alleviate negative affect. This effect is consistent with previous literature; it may well be that binge eating distracts one from one’s experience of subjective distress (Smith et al., 2007, p. 195).
DISTINGUISHING BETWEEN NEGATIVE & POSITIVE REINFORCEMENT IN EATING DISORDERS
While expectancy theory does state that both forms of reinforcement contribute to the risk process of developing an eating disorder, it is unclear whether they predict different forms of disordered eating. In an elegant manner, researchers demonstrated that in a sample of middle school girls, the relief-from-distress (i.e., negative reinforcement) eating expectancy predicted higher levels of binge eating behaviour over the following year but did not predict higher levels of a less maladaptive form of overeating: “eating too much in social or celebratory contexts” (Combs et al., 2010).
Contrastingly, the expectancy that eating results in fun and pleasure (i.e., positive reinforcement) did predict higher levels of social/celebratory overeating, but it did not predict higher levels of binge eating. Accordingly,
These differential predictions support the theoretical distinction between the two expectancy constructs with respect to their role in the risk process (Combs et al., 2010, p. 30).
Of course, it is important to bear in mind that this study was the first to differentiate between the two different forms of learning about eating across time. Consequently, support for this theoretical distinction is still preliminary and we cannot jump to conclusions just yet. I do think that the results are rather promising and am looking forward to reading up on related studies in the future.
EATING EXPECTANCIES AND ANOREXIA
My search for eating expectancies in patients with AN and/or in relation with pathological restriction was unfortunately unfruitful. Taking into account the fact that many patients with AN binge and purge as well, eating expectancies related to positive/negative reinforcement could apply to this population (perhaps to a different degree). Fortunately, Fitzsimmons-Craft et al. (2013) included AN patients in their sample.
EATING EXPECTANCIES IN RELATION TO RECOVERY
Fitzsimmons-Craft et al. wanted to compare and contrast eating expectancies of healthy controls and former eating disorder patients (all females) at varying stages of recovery (i.e., full recovery, partial recovery, or active eating disorder). Participants were grouped on the basis of their stage of recovery rather than their ED diagnoses. Recovery was defined according to physical, behavioural, and psychological indices.
Participants ranged from 16 to 40 years of age, with the majority of the sample consisting of Caucasians (91.6%). Of the 88 with a history of an ED, 55 were ill (77% EDNOS, 17% AN, 6% BN), 15 were partially recovered, and 20 were fully recovered. There were 67 healthy controls.
To be considered fully recovered, an individual would have to have a BMI of at least 18.5, reported no binge eating, purging, or fasting in the last 3 months, and score within the normal range on the Eating Disorder Examination scale (which measures food- and weight-related psychopathology/behaviours).
To be considered partially recovered, an individual would have to meet all of the criteria of full recovery except for psychology recovery.
RESULTS AND DISCUSSION
1. Fully recovered participants resembled healthy controls. Both had:
- Low expectancy that eating helps manage negative affect (a type of negative reinforcement expectancy);
- High expectancy that eating is pleasurable and useful as a reward (a type of positive reinforcement expectancy);
- Low expectancy that eating leads to feeling out of control.
The authors explain that:
These results are encouraging given that research has indicated that addressing cognitive biases, such as maladaptive expectations regarding eating, is an important part of working toward recovery from an eating disorder (p. 1045).
2. Conversely, those who were partially recovered resembled those who had an active eating disorder.
Recall that partial recovery differed from full recovery in that individuals still grappled with psychological aspects of the behaviour. This implies that the endorsement of such expectancies (high expectancy that eating helps manage negative affect and leads to feeling out of control, and low expectancy that eating is pleasurable) is associated with a lack of psychological recovery, despite having recovered physically and behaviourally.
Furthermore, fully recovered individuals were less likely than those with an active ED to expect eating to alleviate negative affect and more likely to find eating to be pleasurable. Also, compared to partially recovered and active ED groups, fully recovered individuals experienced lower levels of the expectancy that eating would lead to out of control feelings.
Partially recovered individuals did not differ from those with an active ED with regards to their expectancy that eating leads to feeling out of control. As explained by the investigators, this finding have two implications:
- Despite being behaviourally and physically healthy, this expectancy appears to have remained in partially recovered individuals. This may be manifested among some people as subjective binge eating (i.e., “out-of-control eating episodes that are not objectively large”).
- As the only expectancy that differentiated fully recovered individuals from the partially recovered and active eating disorder groups, it may be especially worthwhile to target the assumption that eating would lead to feeling out of control.
We can tentatively gather from this study that firstly, different stages of recovery can be marked by varying eating expectancies. Secondly, eating expectancies can be of clinical utility, especially when comprehensively evaluating a person’s stage of recovery. In addition, the expectancy that eating would lead to feeling out of control could be targeted during treatment.
However, because of the study’s design, we are unable to tell if these expectancies were presented before the onset of the disorder OR if the disorder led to the development of these expectancies. Likewise, we cannot pinpoint the specific, temporal role the expectancies play in achieving full recovery (e.g., a prerequisite to full recovery or the consequence of).
Taking into account the study’s findings as well as those of previous studies, eating expectancies — as a construct — should be explored further because it has potential to enhance therapeutic outcomes and understanding.
FINAL NOTES: DIRECTIONS FOR FUTURE RESEARCH
You’ve probably noticed that I’ve only discussed eating expectancies. I did not delve into dieting and thinness expectancies. This is mainly due to the fact that Fitzsimmon-Craft et al.’s paper solely focuses on eating expectancies. It is unlikely that thinness, dieting, and eating expectancies are unrelated (see here and here). Thus, future studies investigating expectancies in relation to recovery should cover dieting, thinness, AND eating expectancies.
While Fitzsimmons-Craft et al. (2013) did include those with AN (either recovered or actively engaged) in their study’s sample, no analyses were conducted to compare expectancies across diagnostic groups. Thereby, future studies could explore expectancy theory solely with regards to anorexia, which would permit us to gain a specialized wealth of information.
The Eating Expectancy Inventory is one of the most commonly used questionnaires in this field. But if we’re investigating eating expectancies in a transdiagnostic sample, I believe that the inventory’s subscales do not adequately capture restriction-related expectancies . It is, in my opinion, slightly biased towards bulimic presentations. Because I could not find any studies examining eating expectancies in AN, I am going off on what I’ve read and learnt about the disorder in general. Restriction serves a purpose, albeit a dysfunctional one. Patients often report that restriction helps with numbing difficult emotions and/or that restriction relieves anxiety (read Tetyana’s post here). Moreover, restriction is not limited to those with AN; it is also present in many (all?) BN patients. For instance, Bruce et al. (2009) demonstrated that BN groups had higher thinness and dietary restriction expectancies than controls. Hence, and in line with my first point, a scale like The Thinness and Restriction Expectancy Inventory should be included when measuring eating expectancies in those with AN and BN.
Considering that the studies by Combs et al. (2010) and Smith et al. (2007) have shown that binge eating led to stronger endorsement of the expectancy that eating helps alleviate distress (i.e., early-onset binge eating could further heighten risk), it may be important to intervene with early-onset binge eating. Studies should be conducted to see if early intervention with regards to negative reinforcement eating expectancies lead to improved outcomes.
And, finally, all aforementioned studies suffer from the same limitations: lack of gender and ethnic/racial diversity in their sample populations. So it is hard to say how these findings generalize to other subgroups.
Bohon C, Stice E, & Burton E (2009). Maintenance factors for persistence of bulimic pathology: a prospective natural history study. International Journal of Eating Disorders, 42 (2), 173-8 PMID: 18951457
EE&rft.au=Keatts+DA&rft.au=Bardone-Cone+AM&rfe_dat=bpr3.included=1;bpr3.tags=Medicine%2CPsychology%2CNeuroscience%2CPsychiatry%2C+Eating+Disorders%2C+Eating+Expectancies%2C+Neuroscience%2C+Psychology%2C+Behavioral+Neuroscience%2C+Anorexia+Nervosa%2C+Eating">Fitzsimmons-Craft EE, Keatts DA, & Bardone-Cone AM (2013). Eating Expectancies in Relation to Eating Disorder Recovery. Cognitive Therapy and Research, 37 (5) PMID: 24089581
Smith GT, Simmons JR, Flory K, Annus AM, & Hill KK (2007). Thinness and eating expectancies predict subsequent binge-eating and purging behavior among adolescent girls. Journal of abnormal psychology, 116 (1), 188-97 PMID: 17324029